fbinary-methods            package:msbase            R Documentation

_B_i_n_a_r_y _m_e_a_s_u_r_e_s _f_o_r _p_a_i_r_w_i_s_e _p_e_a_k-_l_i_s_t _c_o_m_p_a_r_i_s_o_n

_D_e_s_c_r_i_p_t_i_o_n:

     Binary measures are comptued on entries of a contingency table.
      Implements the:

        *  Relative Mutual information,

        *  Fowlkes Mallows statistics M_11/sqrt(M_1Y*M_1X),

        *  Gower coefficient
           (M_01^XY+M_10^XY)/(M_01^XY+M_10^XY+M_11^XY),

        *  Huberts Gamma

_A_r_g_u_m_e_n_t_s:

     obx: see below in *Methods* section

     oby: see above in *Methods* section

   error: measurement error

     ppm: if 'TRUE' then error in parts per million(ppm), in arbitrary
          units otherwise.

  weight: should mass accuracy be weighted

    uniq: if 'TRUE' compute non-crossing matching.

  method: type of dissimilarity:

             *  rmi - relative mutual information

             *  fm - fowlkes mallows statistics

             *  hg - huberts gamma

             *  gower - gower coefficient

       N: default 0 - total length of alinged peak-lists.

   range: experimental

_M_e_t_h_o_d_s:

     _o_b_x = "_n_u_m_e_r_i_c", _o_b_y = "_n_u_m_e_r_i_c" expects two numeric vectors as
          parameters, returns scalar

     _o_b_x = "_M_a_s_s_v_e_c_t_o_r", _o_b_y = "_M_a_s_s_v_e_c_t_o_r" returns scalar ('numeric')

     _o_b_x = "_M_a_s_s_v_e_c_t_o_r_l_i_s_t", _o_b_y = "_M_a_s_s_v_e_c_t_o_r" returns a vector with
          all pairwise dissimilaritities.

     _o_b_x = "_M_a_s_s_v_e_c_t_o_r_l_i_s_t", _o_b_y = "_N_U_L_L" returns object of class
          'dist'

     _o_b_x = "_l_i_s_t", _o_b_y = "_N_U_L_L" returns object of class 'dist'

_A_u_t_h_o_r(_s):

     Witold E. Wolski witek96@users.sourceforge.net <URL:
     http://r4proteomics.sourceforge.net>

_E_x_a_m_p_l_e_s:

     # resolve multiple matches.
     data(pldata)
     pl1 <- pldata[[1]]
     pl2 <- pldata[[2]]
     fbinary(pl1,pl2,error=400,ppm=TRUE,theta=1,weight=FALSE,method="rmi",uniq=TRUE,N=0)
     # dont resolve multiple matches.
     fbinary(pl1,pl2,error=400,ppm=TRUE,theta=0.2,weight=FALSE,method="rmi",uniq=FALSE)
     fbinary(pl1,pl2,error=400,ppm=TRUE,theta=0.2,weight=FALSE,method="hg")
     fbinary(pl1,pl2,error=400,ppm=TRUE,theta=0.2,weight=TRUE,method="gower")
     fbinary(pl1,pl2,error=400,ppm=TRUE,theta=0.2,weight=TRUE,method="fm")
     # seach with one peak-list in a list of peak-lists.
     fbinary(pldata,pl1,error=400,ppm=TRUE,theta=0.2,weight=TRUE,method="fm")
     # compute distances (dissimilarities) and cluster.
     tmp <- fbinary(pldata,NULL,error=400,ppm=TRUE,theta=0.2,weight=TRUE,method="fm")
     plot(hclust(tmp,method="average"))

